995 PHASE 3 STUDY EVALUATING THE EFFICACY AND SAFETY OF VONOPRAZAN 20 MG VERSUS LANSOPRAZOLE 30 MG IN THE TREATMENT OF ENDOSCOPICALLY CONFIRMED DUODENAL ULCERS IN ASIAN SUBJECTS WITH OR WITHOUT HELICOBACTER PYLORI INFECTION

Background:Although previous studies had demonstrated the seasonal variations of upper gastrointestinal bleeding (UGIB), these studies failed to address data autocorrelations as well as factors that would modulate the incidences of UGIB including the use of proton pump inhibitors and H pylori eradication.No studies have examined the seasonal variations of lower GIB (LGIB).Aim: To investigate the seasonal variations and the association with various meteorological parameters of the incidences of UGIB and LGIB in Hong Kong over the past 10 years.Methods: This is a population-based study based on the public healthcare database of Hong Kong.We included all patients hospitalized for UGIB and LGIB and all prescription records in the public hospitals of Hong Kong between 2009 and 2018.Meteorological parameters of the same period were obtained from Hong Kong Observatory.The monthly age-and sex-standardized GIB data were fitted to monthly meteorological data including average temperature (AT), maximum temperature (MaxT), minimum temperature (MinT), temperature range (TR), average precipitation (Prep), average atmospheric pressure (AtomP) and average relative humidity (RH), at different month-lags and adjusted for total aspirin, PPI prescriptions and number of patients received HP eradication using the autoregressive integrated moving average (ARIMA) model.Results: A total of 53,911 UGIB and 67,276 LGIB events were detected over the 10-year period.The monthly UGIB incidences showed an obvious seasonal pattern (after removing the long-term trend using additive decomposition model) while no obvious seasonal pattern was observed for LGIB ( Figure 1).The median incidences of UGIB were higher in winter months (Dec to Mar) than in summer months (Jun to Sep, Wilcoxon P = 0.005).The correlation coefficients showed that the seasonal components of AT, MaxT, MinT, TR, Prep and AtomP correlated with UGIB (absolute values >0.8), but not LGIB (all <0.40).In the ARIMA model fitted to UGIB, but not LGIB, a significant 12-month seasonal pattern was detected.After adjusting for aspirin, PPI and H. pylori eradication, we found that the UGIB incidence was inversely associated with AT, MaxT and MinT at lag 0 and lag 1 (month), but positively associated with TR at lag 1 and AtomP at lag 0 and lag 1 ( Figure 2A).Although there was no seasonal variation of LGIB, temperature parameters including AT, Max T, Min T and TR were significantly associated with the LGIB incidences at lag 0, after adjusting for aspirin, but the impacts were less than UGIB (Figure 2B).Conclusion: UGIB showed a significant seasonal pattern with higher incidences in winter months, which was associated with air temperature and indirectly, the associated atmospheric pressure.Although LGIB did not show obvious seasonal pattern, air temperature also impacted the incidences of LGIB to a lesser extent.